EP1263180A1 - Decision feedback equalizer whereby feedforward filter includes post-cursor and pre-cursor taps - Google Patents
Decision feedback equalizer whereby feedforward filter includes post-cursor and pre-cursor taps Download PDFInfo
- Publication number
- EP1263180A1 EP1263180A1 EP02010714A EP02010714A EP1263180A1 EP 1263180 A1 EP1263180 A1 EP 1263180A1 EP 02010714 A EP02010714 A EP 02010714A EP 02010714 A EP02010714 A EP 02010714A EP 1263180 A1 EP1263180 A1 EP 1263180A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- taps
- filter
- cursor
- feed forward
- equalizer according
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L25/00—Baseband systems
- H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
- H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
- H04L25/03006—Arrangements for removing intersymbol interference
- H04L25/03012—Arrangements for removing intersymbol interference operating in the time domain
- H04L25/03019—Arrangements for removing intersymbol interference operating in the time domain adaptive, i.e. capable of adjustment during data reception
- H04L25/03057—Arrangements for removing intersymbol interference operating in the time domain adaptive, i.e. capable of adjustment during data reception with a recursive structure
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L25/00—Baseband systems
- H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
- H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
- H04L25/03006—Arrangements for removing intersymbol interference
- H04L2025/03433—Arrangements for removing intersymbol interference characterised by equaliser structure
- H04L2025/03439—Fixed structures
- H04L2025/03445—Time domain
- H04L2025/03471—Tapped delay lines
- H04L2025/03484—Tapped delay lines time-recursive
- H04L2025/0349—Tapped delay lines time-recursive as a feedback filter
Definitions
- the present invention relates to equalizer devices and concerns in particular equalizers with feedback of decisions of the type currently called decision-feedback equalizer (or, briefly, DFE).
- an equalizer of this type essentially includes two filters 10 and 12 acting as feed forward filter and feedback filter, respectively.
- the signal destined to be equalized and submitted to detection is received on an input line indicated with 10a.
- the subject signal is applied to a first input 14a of a sum node 14.
- the signal produced in the sum node 14 is brought to the input of a device 16 performing a symbol-by-symbol detection function and produces at its output the sequence of detected symbols, on a line indicated with 18. This last sequence is newly sent towards the feedback filter 12 whose output is sent to the other input of the sum node 14 indicated with 14b.
- both the filters 10 and 12 are equipped with taps spaced by a time slot equal to the symbol interval.
- the filter 10 is essentially a linear transverse equalizer susceptible to assume the shape of an adapted whitened filter - in the case of a DFE equalizer of the 0 forcing (zero-forcing) type in which both the filters 10 and 12 are of the impulse infinite response (IIR) type -.
- the operation of the equalizer grounds on the fact that the feedback filter 12 (having as input the sequence of decisions on the symbols previously detected) is used to eliminate the inter-symbolic interference portion caused by the symbols previously detected from the current estimate.
- the optimization of the two filters, each one having a finite number of coefficients, is generally made according to the criterion based on the minimization of the mean square error (MSE).
- the (estimated) signal Î k relevant to the k-th symbol present at the output of the sum node 14 of figure 1 can be generally expressed according to a relation of the following type: were v k-j indicates the sequence of signals at the input of filter 10 and are the symbols previously detected while c j are the weighting coefficients associated to the different filter taps, that is K 1 +1 taps in the transversal filter 10 and K 2 taps in the feedback filter 12.
- the equalizer is intrinsically non-linear, due to the fact that the feedback filter 12 acts on the basis of the symbols previously detected so to keep into account the previous signal "history".
- Scope of the present invention is to implement an equalizer of the type specified above, additionally perfected concerning its performances, mainly in relation to the possible application for the identification and equalization of the transmission channel in the frame of the connections affected by selective fading in frequency.
- the solution according to the invention foresees to employ, in the frame of a structure of decision-feedback equalizer of the type previously described, a structure in which the signal post-cursors are considered in addition to the pre-cursors in the feed forward filter (indicated with 10 in the description already made with reference to figure 1).
- the feed forward filter is implemented as a transversal filter with M1 pre-cursors and M2 post-cursors preferably implemented with sampling frequency equal to twice the symbol frequency.
- the Applicant has ground to think that the surprising performance improvement that can be obtained with the invention is tied to the fact that the adoption of a structure considering also the post cursors for the realization of the feed forward filter is such to give the equalizer the characteristics of a non-minimum phase system, thus improving the effectiveness of the compensation action of noise phenomena of the transmission channel that can be configured they too as non-minimum phase phenomena.
- the transmission channel marked with C in figure 2 supplies at output the signal x(t) on which the equalization operation and detection of symbols transmitted has to be carried out.
- the above-mentioned signal x(t) is submitted to sampling in a sampling device 20 (of the type known in itself) however operating with a sampling frequency equal to twice the symbol frequency.
- the sampled signal brought to the input of the feed forward filter 10 consists of a sequence of samples separated each other by a time interval equal to Ts/2.
- the numeric reference 22 indicates a down sampling unit performing the elimination of a sample over two, among those coming out from filter 10.
- a flow of samples reduced to one half and therefore susceptible of being brought to the first input 14a of the sum node 14 with a number of values in the time unit equal to the homologue number of values simultaneously brought to the second input 14b starting from the output of the feedback filter 12.
- the estimated set is simply obtained through the convolution process of the signal consisting of the samples received at Ts/2 (remember that the symbol interval is herein indicated with Ts) and of the preamble with the impulsive response of the adaptive filter.
- the switch 24 is set on the position shown with a continuous line in figure 2, causing that the sequence of estimated symbols supplied on the output line 18 of the equalizer 1 is sent back also towards the input of the feedback filter 12.
- the feedback filter 12 coefficients are indicated with Pi while in the same relation the transversal filter 10 coefficients have been indicated with Ci.
- the described configuration includes three post-cursors (weighted with coefficients C1, C2 and C3), a pre-cursor (weighted with the coefficient C-1) and the central tap concerning the current term x(2n) (weighed with the Co coefficient).
- the table shows the results in terms of S/MSE ratio, obtained for sixteen different equalizer structures based on the use of the number of pre-cursors (M1), post-cursors (M2) and coefficients of the feedback filter (L) indicated in the column on the far left of the table.
- Data indicated in the first and in the last line of the table refer to equalizers in which the transversal filter 10 has been designed with sampling frequency equal to the symbol frequency.
- the solution according to the invention can be perfectly used for the possible implementation of the equalizer through DSP and this just in consideration of the possibility to contain the computational load tied to the implementation of the equalizer within acceptable values.
Abstract
A decision-feedback equalizer, including a transversal filter (10) fed with a signal
containing symbols to submit to detection. The equalizer includes also a feedback
filter (12) fed with the sequence of symbols already submitted to detection as well as
a detection element (16) performing the detection action starting from the signal
processed by the transversal filter (10) and by its sequence of symbols already
submitted to feedback adapted by the feedback filter (12). The transversal filter (10)
includes both pre-cursor taps and post-cursor taps. Preferably, post-cursor
taps are at least two or three, thus being in number higher than or equal to the
pre-cursor taps and to the feedback filter taps (12). In view of the reduced
number of necessary taps, the equalizer can be easily implemented through a
DSP.
Description
The present invention relates to equalizer devices and concerns in particular
equalizers with feedback of decisions of the type currently called decision-feedback
equalizer (or, briefly, DFE).
According to the classical structure (taken as pattern for the preamble of claim 1
and schematically recalled in figure 1 of the attached drawings), an equalizer of this
type essentially includes two filters 10 and 12 acting as feed forward filter and feedback
filter, respectively.
The signal destined to be equalized and submitted to detection is received on an
input line indicated with 10a. After processing in filter 10, the subject signal is applied to
a first input 14a of a sum node 14. The signal produced in the sum node 14 is brought
to the input of a device 16 performing a symbol-by-symbol detection function and
produces at its output the sequence of detected symbols, on a line indicated with 18.
This last sequence is newly sent towards the feedback filter 12 whose output is sent to
the other input of the sum node 14 indicated with 14b.
In general, both the filters 10 and 12 are equipped with taps spaced by a time slot
equal to the symbol interval. The filter 10 is essentially a linear transverse equalizer
susceptible to assume the shape of an adapted whitened filter - in the case of a DFE
equalizer of the 0 forcing (zero-forcing) type in which both the filters 10 and 12 are of
the impulse infinite response (IIR) type -.
Actually, the operation of the equalizer grounds on the fact that the feedback filter
12 (having as input the sequence of decisions on the symbols previously detected) is
used to eliminate the inter-symbolic interference portion caused by the symbols
previously detected from the current estimate. The optimization of the two filters, each
one having a finite number of coefficients, is generally made according to the criterion
based on the minimization of the mean square error (MSE).
All that indicated above, and in particular the criteria enabling to determine all the
coefficients of filters 10 and 12 using the minimization of the mean square error,
corresponds to criteria already well known in the technique.
In particular, for a general processing in the decision-feedback equalization
technique, useful reference can be made to John G. Proakis volume - "Digital
communications" McGraw Hill, 4th edition, and in particular to section 10.3 on page 638
and subs.
Concerning the calculation of the coefficients of filters 10 and 12 carried out with
equations in the matricial form (Wiener-Hopf equations, through determination of the
relevant Hermitiana equivalent autocorrelation matrix that can be expressed through
Cholesky factorization) reference can be made, in addition to the already mentioned
Proakis volume, to other reference papers such as:
- M. Bellanger, "Digital Processing of Signals", Jon Wiley & Sons,
- Fuzhen Zhang, "Matrix Theory", Springer Verlag, and
- S. Benedetto, E. Biglieri, V. Castellani, "Theory of digital transmission", Jackson Libri.
Whichever are the specific implementation details adopted, all the "classical"
dealings of the decision-feedback equalizer, and in particular those previously
considered, foresee the resort to a structure based on the utilization of the sole so-called
pre-cursors of the signal for the implementation of the transversal filter 10,.
In other words, having recourse to the expression (10.3-1) given on page 638 of
the already mentioned Proakis book, the (estimated) signal Îk relevant to the k-th
symbol present at the output of the sum node 14 of figure 1 can be generally
expressed according to a relation of the following type:
were vk-j indicates the sequence of signals at the input of filter 10 and are the
symbols previously detected while cj are the weighting coefficients associated to the
different filter taps, that is K1+1 taps in the transversal filter 10 and K2 taps in the
feedback filter 12.
It could be appreciated that the equalizer is intrinsically non-linear, due to the fact
that the feedback filter 12 acts on the basis of the symbols previously detected so to
keep into account the previous signal "history".
Scope of the present invention is to implement an equalizer of the type specified
above, additionally perfected concerning its performances, mainly in relation to the
possible application for the identification and equalization of the transmission channel
in the frame of the connections affected by selective fading in frequency.
According to the present invention, this scope is reached thanks to an equalizer
device having the characteristics specifically described in the following claims.
In practice, the solution according to the invention foresees to employ, in the
frame of a structure of decision-feedback equalizer of the type previously described, a
structure in which the signal post-cursors are considered in addition to the pre-cursors
in the feed forward filter (indicated with 10 in the description already made with
reference to figure 1).
Preferably, the feed forward filter is implemented as a transversal filter with M1
pre-cursors and M2 post-cursors preferably implemented with sampling frequency
equal to twice the symbol frequency.
The experiences conducted by the Applicant have demonstrated that -
surprisingly - the fact to use also the post-cursors in the implementation of the feed
forward filter enables to obtain unexpected improvements in terms of ratio at the
signal/mean square error (S/MSE).
All the above, through implementation of a structure that proves to be largely
insensible to possible errors (jitter) in the development of the detected signal sampling
operation and with the additional possibility to obtain an important performance
improvement keeping the total number of taps unchanged (that is, reducing the number
of pre-cursors and/or taps in the feedback filter by a value corresponding to the number
of post-cursors inserted in the feed forward filter) or even reducing the total number of
the taps (that is, reducing the number of pre-cursors and/or taps in the feedback filter
by an amount higher than the number of post-cursors inserted in the feed forward
filter).
This aspect results particularly important since it allows restricting, and in some
cases also reducing, the computational burden tied to the realization of the different
taps, with the consequent possibility to implement the equalizer as a whole through a
DSP.
Though not tying to any specific theory on the matter, the Applicant has ground to
think that the surprising performance improvement that can be obtained with the
invention is tied to the fact that the adoption of a structure considering also the post
cursors for the realization of the feed forward filter is such to give the equalizer the
characteristics of a non-minimum phase system, thus improving the effectiveness of
the compensation action of noise phenomena of the transmission channel that can be
configured they too as non-minimum phase phenomena.
The invention shall be now described, as an example but not limited to the same,
with reference to the attached drawings, in which:
- Figure 1, relevant to the general structure of a digital-feedback equalizer, has already been described above,
- Figure 2 shows the block diagram of the structure of an equalizer according to the invention, and
- Figure 3 shows the block diagram of the structure of one of the elements included in the equalizer of figure 2.
In particular, in the diagram of figure 2 elements identical or equivalent from the
structural point of view to those previously described have been marked with the same
reference numbers already appearing in figure 1.
In the frame of the decision-feedback equalizer shown in figure 2 and globally
indicated with 1 appear basic elements equivalent to those previously described from
the structural point of view, that is:
- a feed forward transversal filter 10 (which, as better described below with reference to figure 3, foresees - in the case of the invention - also the use of the signal post-cursors),
- a detector (decision element) operating symbol by
symbol 16, and - a feedback filter or
feedback 12.
The transmission channel marked with C in figure 2 supplies at output
the signal x(t) on which the equalization operation and detection of symbols transmitted has to be carried out.
the signal x(t) on which the equalization operation and detection of symbols transmitted has to be carried out.
In the currently preferred embodiment of the invention, it is foreseen that the
above-mentioned signal x(t) is submitted to sampling in a sampling device 20 (of the
type known in itself) however operating with a sampling frequency equal to twice the
symbol frequency.
Consequently, indicating with Ts the symbol interval, the sampled signal brought
to the input of the feed forward filter 10 consists of a sequence of samples separated
each other by a time interval equal to Ts/2.
The above with the possible presence of a temporal error in the development of
the sampling action, the value of said error being generally represented by a time slot
t0: in these conditions, t0=0 represents the absence of sampling error, while t0/Ts=0.25
represents the maximum value of said error.
In a symmetrical way, the numeric reference 22 indicates a down sampling unit
performing the elimination of a sample over two, among those coming out from filter 10.
The above in order to generate, starting from the flow of filtered samples at the output
of filter 10, a flow of samples reduced to one half and therefore susceptible of being
brought to the first input 14a of the sum node 14 with a number of values in the time
unit equal to the homologue number of values simultaneously brought to the second
input 14b starting from the output of the feedback filter 12.
In the diagram of figure 2 it was also shown a deviator element 24 that, initially
brought to the position indicated with a dotted line during the starting estimate phase of
the equalizer operation procedure, enables to initialise the set of the equalizer
coefficients.
This result is obtained, for instance, performing (in the known way) the
minimization of the mean square error (MSE) so to minimize the deviation e(n)
between the estimated set (detected at the output of the sum node 14) and a target
signal or preamble s(n) - let as known - brought to an additional input line 26 of the
equalizer 1.
In this first phase, being known in reception the target signal s(n), the estimated
set is simply obtained through the convolution process of the signal consisting of the
samples received at Ts/2 (remember that the symbol interval is herein indicated with
Ts) and of the preamble with the impulsive response of the adaptive filter.
After initialization of the filter coefficients, the switch 24 is set on the position
shown with a continuous line in figure 2, causing that the sequence of estimated
symbols supplied on the output line 18 of the equalizer 1 is sent back also towards the
input of the feedback filter 12.
In these conditions, which correspond to the normal operation of equalizer 1, in
reception it is known only the signal received and the equalizer output is obtained from
the convolution process of the signal made of the samples received at Ts/2 and of the
decisions y(n) with the impulse response of the adaptive filter.
In the preceding relation, the feedback filter 12 coefficients are indicated with Pi
while in the same relation the transversal filter 10 coefficients have been indicated with
Ci.
While the solutions according to the known technique foresee the use of
transversal filters showing only pre-cursors (therefore, filters in which, in the relation
recalled above, M1≠0 and M2=0), the solution according to the invention foresees also
the use of the post-cursors, therefore having in this case also M2≠0.
All the above according to the criteria better shown in the diagram of figure 3,
which refers - for example purposes - to the possible implementation of filter 10 as a
transversal filter including four delay elements 101 and five taps 102 controlled by a
sum node 103. Therefore, the described configuration includes three post-cursors
(weighted with coefficients C1, C2 and C3), a pre-cursor (weighted with the coefficient
C-1) and the central tap concerning the current term x(2n) (weighed with the Co
coefficient). It could be appreciated that, according to the current use, the component
of the output signal y(2n) corresponding to the homologue sample of the input signal
(X(2n)) has been actually considered as incorporated between the pre-cursors (see
also the formula 10.3-1 of Proakis volume, wide reference to which has been made
before).
The experiences made by the Applicant demonstrate that - surprisingly and
unexpectedly - at equal linear taps, that is calculation complexity, the equalizer
decision-feedback in which the filter 10 operates both on pre-cursors and on post-cursors
shows better performances in terms of signal/mean square error (S/MSE) ratio
compared to decision-feedback equalizers with pre-cursors only. All the above mainly
in presence of C channels affected by selective fading in frequency and in
independently from the best sampling phase.
These results are indicated in the following table.
M1 M2 L | Bc=0dB fn=0Hz t0=0 | Bc=0dB fn=0Hz t0/Ts=0.25 | f_min Bc=30dB fn=5MHz t0=0 | f_min Bc=30dB fn= 5Mhz t0/Ts=0.25 | f_no_min Bc=30dB fn= 5Mhz t0=0 | f_no_min Bc=30dB fn=5MHz t0/Ts=0.25 |
5 5 0 | 63.0 dB | 24.6 dB | 11.8 dB | 11.6 dB | 12.1 dB | 12.2 |
5 5 0 | 64.3 | 63.5 | 11.4 | 10.9 | 11.5 | 11.5 |
2 2 0 | 58.9 | 48.3 | 10.8 | 10.8 | 11.0 | 11.0 |
3 2 0 | 58.3 | 56.6 | 10.9 | 10.9 | 11.0 | 11.0 |
2 3 0 | 59.0 | 59.0 | 11.0 | 10.9 | 11.1 | 11.0 |
3 3 0 | 58.7 | 58.7 | 10.8 | 10.8 | 10.8 | 11.0 |
2 2 1 | 58.9 | 54.4 | 54.1 | 55.4 | 31.9 | 22.1 |
3 2 1 | 58.3 | 58.7 | 53.8 | 55.2 | 30.1 | 23.4 |
2 2 2 | 59.2 | 57.5 | 53.9 | 51.7 | 42.7 | 24.6 |
3 2 2 | 58.6 | 58.8 | 54.2 | 52.0 | 42.1 | 29.7 |
2 3 1 | 59.0 | 56.3 | 54.1 | 51.6 | 50.7 | 51.9 |
3 3 1 | 58.7 | 58.6 | 54.5 | 51.9 | 49.4 | 51.9 |
2 3 2 | 52.4 | 57.7 | 54.1 | 51.4 | 49.3 | 51.5 |
3 3 2 | 53.5 | 58.6 | 54.6 | 51.7 | 46.8 | 52.1 |
5 0 5 | 63.0 | 46.7 | 45.1 | 33.1 | 19.5 | 13.0 |
5 0 5 | 63.0 | 22.8 | 35.6 | 21.7 | 11.8 | 12.2 |
The table shows the results in terms of S/MSE ratio, obtained for sixteen different
equalizer structures based on the use of the number of pre-cursors (M1), post-cursors
(M2) and coefficients of the feedback filter (L) indicated in the column on the far left of
the table.
The results obtained are ordered in six columns respectively referring to the
following channel conditions:
- absence of selective fading in frequency (Bc=0 dB, fn=0Hz) and absence of sampling errors (t0=0),
- absence of selective fading in frequency and maximum sampling error (t0/Ts=0.25),
- presence of minimum phase fading (f_min), with 30 dB notch depth (Bc) and 5 MHz notch frequency (fn), and t0=0,
- presence of minimum phase fading as in the previous case, but with maximum sampling error t0/Ts=0.25,
- presence of non-minimum phase fading (f_no_min), with 30 dB notch depth and 5MHz notch frequency, and t0=0, and
- presence of non-minimum phase fading, as in the previous case, and of maximum sampling error t0/Ts=0.25.
Data indicated in the first and in the last line of the table refer to equalizers in
which the transversal filter 10 has been designed with sampling frequency equal to the
symbol frequency.
From the evaluation of the subject table it can be noticed that in lack of selective
fading in frequency all the considered equalizers show - even with reading instant t0
different from the best one - a good performance, with a rather high S/MSE ratio.
In presence of minimum phase selective fading in frequency with notch depth
and notch frequency in the terms previously recalled, the performance of linear
equalizers reveals to be poor (S/MSE ratio 10-11 dB) even with reading instant t0 equal
to the best one. In these conditions a substantial worsening of the performances of the
traditional DFE equalizers can be noticed (see the last two lines of the table,
concerning equalizers with ten taps evenly shared between the feed forward filter 10
and the feedback filter 12).
The results indicated in the columns on the far right demonstrate that degrading
of the performances of linear and DFE equalizers according to the known technique,
becomes more evident in presence of non-minimum phase fading, in particular when
we have a reading instant different from the best one (column on the far right of the
table with t0/Ts=0.25).
On the contrary, it can be noticed that also in these very critical conditions, the
performances of the equalizers according to the invention remain however satisfactory
even in presence of a reduced number of taps (for instance five taps with M1=2, M2=2
and L=1).
Observing the results indicated in the lines seven to fourteen of the attached
table, it might also be noticed that the simple increase by one unit of the post-cursor
number (with the change of M2 from value 2 to value 3) enables to obtain significant
improvements of the equalizer performances.
In particular, while the increase of M2 from value 2 to value 3 involves some
betterment of the S/MSE ratio in the range of 20-30 dB, the increase by one unit of M1
and/or L1 leads to globally limited variations of the performances.
In particular, in exceptionally critical conditions evidenced by the results in the far
right columns of the table, the results obtained with the equalizer in which M1=2, M2=3
and L=1 (therefore with a total number of six taps) result neatly higher than the results
that can be obtained also with a higher number of taps (for instance ten taps) when
post-cursors are not present (M2=0, see the last two lines of the table).
Nor the above-mentioned best result turns out to be significantly improved or
improvable by increasing the number of taps when this increase occurs with an
increase of M1 or L.
The solution according to the invention can be perfectly used for the possible
implementation of the equalizer through DSP and this just in consideration of the
possibility to contain the computational load tied to the implementation of the equalizer
within acceptable values.
It is evident that variants and modifications are possible for those skilled in the art
without departing from the protections of the following claims.
Claims (14)
- Decision-feedback equalizer including a feed forward filter (10) to which its is sent an input signal with symbols to submit to detection, a feedback filter (12) fed with the sequence of the symbols already submitted to detection and a detection element (16) susceptible to perform the detection action of said symbols starting from said input signal processed by said feed forward filter (10) and by the sequence of said symbols already submitted to detection processed by said feedback filter (12), characterized in that said feed forward filter (10) is a transversal filter including at least one post-cursor tap (M2).
- Equalizer according to claim 1, characterized in that it includes a sampling stage (20) in order to submit said input signal to sampling with a sampling time lower than the symbol time (Ts) of said input signal.
- Equalizer according to claim 2, characterized in that said sampling stage (20) operates with sampling time, equal to one half of said symbol time (Ts).
- Equalizer according to claim 2 or claim 3, characterized in that it includes, interposed between said feed forward filter (10) and said detection element (16), a down sampler stage (22) to generate, starting from the sample flow generated by said feed forward filter (10), a reduced flow of samples separated by a time equal to said symbol time (Ts).
- Equalizer according to claim 1, characterized in that in said feed forward filter (10), the number of post-cursor taps (M2) is equal at least to the number of pre-cursor taps (M1).
- Equalizer according to claim 5, characterized in that the number of said post-cursor taps (M2) is higher than the number of said pre-cursor taps (M1).
- Equalizer according to any claim 1, 5 or 6, characterized in that said feedback filter (12) includes a relevant number of taps (L) and in the fact that the number of said post-cursor taps (M2) is equal at least to said relevant number of taps (L).
- Equalizer according to claim 7, characterized in that the number of said post-cursor taps (M2) is higher than said relevant number of taps (L).
- Equalizer according to any claim 1 through 8, characterized in that said feed forward filter includes at least two post-cursor taps.
- Equalizer according to claim 9, characterized in that said feed forward filter (10) includes at least three post-cursor taps.
- Equalizer according to any of the previous claims, characterized in that the total number of the taps of said feed forward filter (10) and of said feedback filter (12) is lower than or equal to ten.
- Equalizer according to any of the previous claims, characterized in that the total number of the taps of said feed forward filter (10) and of said feedback filter (12) is lower than or equal to seven.
- Equalizer according to any of the previous claims, characterized in that the total number of the taps of said feed forward filter (10) and of said feedback filter (12) is lower than or equal to six.
- Equalizer according to any of the previous claims, implemented as a DSP.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
ITMI20011142 | 2001-05-30 | ||
IT2001MI001142A ITMI20011142A1 (en) | 2001-05-30 | 2001-05-30 | EQUALIZER DEVICE |
Publications (1)
Publication Number | Publication Date |
---|---|
EP1263180A1 true EP1263180A1 (en) | 2002-12-04 |
Family
ID=11447777
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
EP02010714A Withdrawn EP1263180A1 (en) | 2001-05-30 | 2002-05-14 | Decision feedback equalizer whereby feedforward filter includes post-cursor and pre-cursor taps |
Country Status (2)
Country | Link |
---|---|
EP (1) | EP1263180A1 (en) |
IT (1) | ITMI20011142A1 (en) |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5513216A (en) * | 1994-10-13 | 1996-04-30 | At&T Corp. | Hybrid equalizer arrangement for use in data communications equipment |
US5561687A (en) * | 1993-12-20 | 1996-10-01 | Adtran | Decision feedback equalizer employing fixed ratio postcursor taps for minimizing noise and intersymbol interference in signals conveyed over high speed data service loop |
-
2001
- 2001-05-30 IT IT2001MI001142A patent/ITMI20011142A1/en unknown
-
2002
- 2002-05-14 EP EP02010714A patent/EP1263180A1/en not_active Withdrawn
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5561687A (en) * | 1993-12-20 | 1996-10-01 | Adtran | Decision feedback equalizer employing fixed ratio postcursor taps for minimizing noise and intersymbol interference in signals conveyed over high speed data service loop |
US5513216A (en) * | 1994-10-13 | 1996-04-30 | At&T Corp. | Hybrid equalizer arrangement for use in data communications equipment |
Also Published As
Publication number | Publication date |
---|---|
ITMI20011142A1 (en) | 2002-11-30 |
ITMI20011142A0 (en) | 2001-05-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US5414733A (en) | Decision feedback equalizer employing fixed ratio postcursor taps for minimizing noise and intersymbol interference in signals conveyed over high speed data service loop | |
US7394849B2 (en) | Decision feedback equalizer with dynamic feedback control | |
US7421021B2 (en) | Adaptive signal equalizer with adaptive error timing and precursor/postcursor configuration control | |
DE60311348T2 (en) | Calculation of the coefficients for a decision-feedback equalizer of variable delay | |
JP2762836B2 (en) | Interference wave canceller | |
KR100447201B1 (en) | Channel equalizer and digital TV receiver using for the same | |
JPH10500830A (en) | Measures against signal interference in digital transmission | |
EP0729254A2 (en) | Fractionally spaced equalizing circuits and method | |
WO2004059899A9 (en) | Adaptive coefficient signal generator for adaptive signal equalizers with fractionally-spaced feedback | |
US20040091041A1 (en) | Adaptive coefficient signal generator for adaptive signal equalizers with fractionally-spaced feedback | |
JP2000507067A (en) | Digital receiver | |
US8937995B2 (en) | Equalizer and equalizing method thereof | |
Pandey et al. | Comparative study of LMS and NLMS algorithms in adaptive equalizer | |
US5524124A (en) | Multiple-filter equalizer for structured digitally modulated signals | |
US7035330B2 (en) | Decision feedback equalizer with dynamic feedback control | |
EP1263180A1 (en) | Decision feedback equalizer whereby feedforward filter includes post-cursor and pre-cursor taps | |
KR100309098B1 (en) | Signal processing method and apparatus, method and apparatus for rejecting cochannel interference signal in modulation/demodulation type digital communication sysyem having reference signal | |
KR100202944B1 (en) | Equalizer | |
US8644369B1 (en) | Equalizer adaptation for heavily compressed or clipped communications signals | |
JP2503715B2 (en) | Adaptive receiver | |
CA1304455C (en) | Adaptive equalizer included in the receiver for a data transmission system | |
US7289559B2 (en) | Method for updating coefficients in decision feedback equalizer | |
KR100698150B1 (en) | Channel equalizer and its method in digital broadcasting receiver | |
JPS59139717A (en) | Automatic equalizer | |
US6442198B1 (en) | Method and apparatus for stabilizing a decision feedback equalizer |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AT BE CH CY DE DK ES FI FR GB GR IE IT LI LU MC NL PT SE TR |
|
AX | Request for extension of the european patent |
Free format text: AL;LT;LV;MK;RO;SI |
|
AKX | Designation fees paid | ||
REG | Reference to a national code |
Ref country code: DE Ref legal event code: 8566 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE APPLICATION IS DEEMED TO BE WITHDRAWN |
|
18D | Application deemed to be withdrawn |
Effective date: 20030605 |